The proposed K23 award will allow Shreya Kangovi, MD, MS an internist and pediatrician to develop expertise in using behavioral science to improve cardiovascular outcomes among patients of low socioeconomic status (SES). Dr. Kangovi and her colleagues developed the IMPaCT intervention, in which community health workers (CHWs) help low-SES patients to improve health behaviors and control CVD risk factors. These patients have challenging lives, and may experience setbacks and failure while attempting the complex process of health behavior change. Interim results from a randomized controlled trial of the intervention suggested that the way patients reacted to failed attempts at health behavior change was an important determinant of subsequent health behavior and ultimately, cardiovascular outcomes. The goal of this K23 proposal is to refine the intervention by helping patients learn from failure, instead of being discouraged by it. The proposed experiments will advance the science of understanding and modifying patients' response to failure. The behavioral science literature describes two ways of responding to failure: learning and avoidance. In learning, individuals use failure signals to improve self-regulation, motivation and performance. In avoidance, individuals become disengaged and perform worse after failure. Two behavioral interventions -positive affect induction (PA) and attribution retraining (AR) -- may promote learning instead of avoidance. PA uses standardized random acts of kindness and self-affirmation to improve emotional resilience after failure. AR teaches individuals to interpret failure as modifiable. PA and AR are proven to improve behavioral outcomes after failure but have not been widely used in healthcare. Dr. Kangovi will work with her mentors and a multidisciplinary team to develop a CHW-delivered positive affect/attribution retraining intervention (CPAAR) for improving patients' response to failure. She will then lead a real-world comparative effectiveness pilot randomized controlled trial (RCT) comparing the effects of the standard versus refined CHW intervention (with CPAAR) on physical activity and CVD outcomes. Dr. Kangovi will be guided by her primary mentor, Dr. Karen Glanz, George A. Weiss University Professor of Epidemiology and Nursing at Penn and Director of the Center for Health Behavior Research and co-mentor, Dr. Judith A. Long, Chief of the Penn Division of General Internal Medicine and an experienced researcher in cardiovascular disparities. This award will provide Dr. Kangovi with skills in translating behavioral science to cardiovascular health promotion interventions. The proposed studies will inform an R01 grant application for a comparative effectiveness RCT of standard versus refined CHW intervention for improving cardiovascular outcomes among low-SES patients.